histogram of selected environmental variable values at the known occurrences - Maxent output using R

103 views
Skip to first unread message

Carin Swart

unread,
Jul 20, 2016, 8:15:55 AM7/20/16
to Maxent
I am interested in using R to produce histograms using presence and background points for selected environmental variables I used in my Maxent model. It was described how to do this in the Maxent tutorial https://www.cs.princeton.edu/~schapire/maxent/tutorial/tutorial.doc page 29 for their Bradypus example but it is not clear how to access the background points file from the Maxent results. It is also not clear how the presence and background points match to the environmental variable values in order to produce the histogram.

Can anyone answer the following:
1. What files do I use in the Maxent output to produce a histogram in R described above?
2. Is there an easy way to convert files from .csv to .swd?

Thanks!

Jamie M. Kass

unread,
Aug 10, 2016, 6:43:48 PM8/10/16
to Maxent
Hey Carin,

Those plots were made using the predictor variable values extracted at the occurrence point locations using the function density() in R, which is sort of like an advanced histogram. If you implement Maxent in R, you will already have your occurrence and background point lat/lon values and your predictor variables as a RasterStack. The next step is using extract() (raster package) to get the raster values at these points, run density() on the variable of choice, and plot it. The vignette for the dismo package walks you through much of this analysis -- please start with this.

Jamie Kass
PhD Candidate
City College, NYC

Tesa M

unread,
Jan 26, 2018, 10:29:19 AM1/26/18
to Maxent
Hello Carin, Jamie,

I am very new to both Maxent and R, and have just run a Maxent model using the GUI interface. What is the best way to get the Maxent output into R so that I can calculate niche overlap? Should I create a RasterStack of the .bil files? I apologize if this is covered elsewhere, and any answers/help is so appreciated!

Tesa

Dimitris Poursanidis

unread,
Jan 26, 2018, 10:32:43 AM1/26/18
to max...@googlegroups.com
Niche overlap - interesting and confusion - depends on the use.

See ENMTools and the related papers on the use of it.

--
You received this message because you are subscribed to the Google Groups "Maxent" group.
To unsubscribe from this group and stop receiving emails from it, send an email to maxent+unsubscribe@googlegroups.com.
To post to this group, send email to max...@googlegroups.com.
Visit this group at https://groups.google.com/group/maxent.
For more options, visit https://groups.google.com/d/optout.

Jamie M. Kass

unread,
Jan 28, 2018, 1:01:03 AM1/28/18
to Maxent
Also please check out the R package ecospat and their vignette. It teaches you how to calculate niche overlap (the Broennimann way).

Jamie Kass
PhD Candidate
City College of NY
Reply all
Reply to author
Forward
0 new messages